Postcolonial Computing:
A Lens on Design and Development
Lilly Irani1, Janet Vertesi1, Paul Dourish1, Kavita Philip2 and Rebecca E. Grinter3
1
Dept. Informatics, 2Dept. Women’s Studies
University of California, Irvine
Irvine, CA 92697
{lirani, jvertesi, jpd, kphilip}@uci.edu
ABSTRACT
As our technologies travel to new cultural contexts and our
designs and methods engage new constituencies, both our
design and analytical practices face significant challenges.
We offer postcolonial computing as an analytical
orientation to better understand these challenges. This
analytic orientation inspires four key shifts in our approach
to HCI4D efforts: generative models of culture,
development as a historical program, uneven economic
relations, and cultural epistemologies. Then, through
reconsideration of the practices of engagement, articulation
and translation in other contexts, we offer designers and
researchers ways of understanding use and design practice
to respond to global connectivity and movement.
Author Keywords
Postcolonial theory, STS, culture, design methods, ICT4D.
ACM Classification Keywords
H5.m. Information interfaces and presentation (e.g., HCI):
Miscellaneous.
INTRODUCTION
Technology travels. It moves around the world in projects
of design and development. Design practice, similarly,
spans continents, both in large-scale processes of
transnational production and in the smaller, local practices
by which technologies are understood and put to use in
different settings. In recent years, HCI has become
especially interested in opportunities surrounding cross-
cultural design practice, with a special focus on HCI design
for “the developing world” (HCI4D). Research in HCI4D
has struggled with a range of complex problems –
technological cultures, digital divides, multiple
stakeholders, economic disparities, and more.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear this notice and the full citation on the first page. To copy otherwise,
or republish, to post on servers or to redistribute to lists, requires prior
specific permission and/or a fee.
CHI 2010, April 10-15, 2010, Atlanta, Georgia, USA.
Copyright 2010 ACM 978-1-60558-246-7/09/04...$5.00.
3
GVU Center and School of Interactive Computing
College of Computing
Georgia Institute of Technology
beki@cc.gatech.edu
In this paper, we examine a series of concerns that are latent
within much research conducted under the umbrella of
HCI4D. We take as our starting point a move from
“development” discourse to postcolonial discourse – that is,
a discourse centered on the questions of power, authority,
legitimacy, participation, and intelligibility in the contexts
of cultural encounter, particularly in the context of
contemporary globalization. Our first goal is to outline the
alternative perspective on HCI4D that we can gain by
looking to the lessons from related disciplines, most
particularly Science and Technology Studies (STS) and
postcolonial studies; we do this with four cases drawn from
the revealing fringes of design-related fieldwork and
history. Our second goal is to show how we might
reconfigure design-oriented cultural encounters in this light.
We label this shift in perspective with the term
“postcolonial computing.”
What is “Postcolonial” Computing?
Formulating the term, we take inspiration from the
articulation of new visions for HCI in programs such as
“mobile computing,” “service-oriented computing,” “urban
computing,” and “ubiquitous computing.” These areas,
however, mark application areas and new forms of
technology. Postcolonial computing is not a new domain or
design space, but an alternative sensibility to the process of
design and analysis. It asserts a series of questions and
concerns inspired by the conditions of postcoloniality but
relevant to any design project – most particularly those in
HCI4D contexts, but in other contexts too.
When we speak of a “postcolonial” approach, we are not
simply focused on the historical conditions of nations and
regions that were once colonies. Postcolonial studies began
with such investigations, but rapidly came to understand
that its topic was actually the historical transformation of
conditions of cultural encounter. Colonial relationships
may have dissolved, and yet the history of global dynamics
of power, wealth, economic strength, and political influence
shape contemporary cultural encounters. For example,
lower cost labor and mineral extraction in Asia and Africa
tacitly undergirds the development of cheaper, faster, and
smaller computers used and sold globally. Colonial tropes
characterizing certain people as in need of enlightenment,
civilization, and development still persist today. These
postcolonial conditions affect China and Sweden as much
as they do India, Britain, Australia, or Kenya. We all live in
a world shaped by colonial histories; we all find ourselves
in postcolonial conditions.
Postcolonial theory has most powerfully demonstrated the
ways projects we engage in for “others” often tell us more
about ourselves. Postcolonial computing, then, is not a
project of making better design for “other” cultures or
places. It is a project of understanding how all design
research and practice is culturally located and power laden,
even if considered fairly general. This specificity is not a
problem to be solved, but a reality that should be central to
design practice – seeing the ways that design is culturally
specific should allow us to broaden the conversation about
what other practices can count as good design.
Postcolonial computing is a shift in perspective motivated
by the challenges and problems of transferring of
technological knowledge, particularly in ICT4D and the
HCI questions it inspires.
THORNY ISSUES IN ICT4D
ICT4D designers face challenges transporting both design
conventions and processes of HCI across cultures. HCI’s
visual conventions have proven not to be universal –
systems effective in the US may fail utterly in Japan or
South Africa. For example, design aesthetics vary wildly
from place to place [21] and taken-for-granted symbolic
literacies, such as recognizing an image representing a GUI
button, are strange in less computer-saturated cultures [23].
The processes of designing and deploying HCI4D across
cultures have proven challenging as well. Researchers
designing for resource-poor but socially interconnected
contexts have proposed shifting from user-centered design
to “communitization” or community-centric design (e.g.
[24]). Community engagement has become important in
requirements elicitation and co-design (e.g. [8,24,28]), as
well as making deployments sustainable [5]. The very
different social, cultural, infrastructural, and economic
situations of HCI4D have required researchers to
substantially adapt HCI methods and practices.
Hardware and connectivity have also produced instructive
case studies of technological failure in international
encounters. Take, for example, the simple matter of a
lightbulb that traveled from Europe to Africa [2]. The
lightbulb’s European designers tightly integrated its
components, hiding the technology in order to user-proof it.
When the lightbulb required adaptation to reach power
sources far away from the room to be lit, the bulb proved
impossible to hack or adapt. The notion of a hermetically
sealed, all-in-one, “plug-and-play” design – seemingly
perfectly adapted to an environment without an extensive
technological infrastructure – turned out, in fact, to render it
useless in the face of local contingencies. Many such well-
intentioned efforts to “migrate” technologies from
industrialized contexts to other parts of the world have
foundered either on infrastructural differences or on social,
cultural, political, or economic assumptions that do not
hold. Such failures of technology transfer led to the rise of
the “Appropriate Technology” (AT) movement in the 1970s
and 1980s [31]. AT focused on fitness for purpose, arguing
that smaller technologies that accounted for local needs,
infrastructures, skills, and materials would be more
effective than large-scale engineering efforts. These
principles, of course, are also central to user-centered
design, and so the emergence of HCI4D as a user-centered
design perspective on cross-cultural development is an
unsurprising development.
Some have sought to predict and understand these problems
of translating HCI knowledge by drawing on taxonomic
models of culture where members of cultural groups are
characterized by traits and averages (e.g. [19]). This model
has been used to explain conflict in organizations,
communication dynamics, and even choices in website
design. However, the assumption that individuals have a
single cultural background is problematic, especially in the
face of contemporary patterns of globalization and
transnationalisms. These models, as Marsden argues [24],
are also of limited help in design because they describe
average tendencies but provide little insight into any
particular person’s cultural experience.
If these models are misleading, then what can we turn to as
a resource for research and design practice? We argue that
STS and postcolonial studies provide understandings of
cultural entanglement and colonial discourses that help us
better understand complex issues of intercultural
engagement around technology design in HCI4D. In the
following section, we illustrate this through explorations of
four connected cases: first, the rhetoric of cultural
differences; second, the problematic rhetoric and practice of
“development”; third, the globalized pattern of economic
relations within which these efforts are embedded; and
fourth, cultural conceptions of knowledge. We then turn to
implications for design practices and methodologies arising
from this analytical turn.
CASE STUDIES
The following cases are drawn from our fieldwork and from
histories of technological travel. These are stories of
connection: of researcher to culturally different users, of
rural Indians to transnational NGOs, of Brazilian engineers
to Apple, and of aboriginal Australians to California. We
examine these mundane experiences in globalized
technology design through the lens of postcolonial
computing.
Reassessing “Culture” Difference
We begin with a case drawn on a research project
conducted by one of us as part of a corporate internship.
The design brief given by the large multinational
corporation hosting my summer internship was clear: the
company was developing a sensor network for the elderly
so that they could remain independent in their homes as
they aged, and wanted to examine the opportunities for
releasing this product in international markets. The target
groups were the BRIC countries (Brazil, Russia, India, and
China), united by their economic status on a global
measurement scale rather than by any cultural similarities.
With little money available for travel, I trekked out into the
mid-sized American town and its suburbs to conduct my
research. There, among immigrant populations, I was told,
I would find examples of these different national cultures,
and could find out from there how their culture conceived
of issues of health and aging.
This design brief may sound familiar to many readers, as
may the conflicts that followed in the ensuing research. The
researcher encountered families that called themselves
“American” in some contexts and “foreign” in others.
These families’ practices bridged communities between
their US locality and international networks, employing a
variety of technical and social resources for what they were,
what they valued, and what they did. Such language
disconnected with the design brief, wherein culture was
taxonomized, reified, and employed as a static
denomination to distinguish between user groups and
communities. This view of culture is quite common as a
way of classifying people, activities, and settings – “Latin”
versus “Asian” cultures, for instance – in terms of
systematic difference between groups. This sense of culture
taxonomized people along lines that might be figured as
geographical, in the sense of regions or nation states, or
collective, as in subcultures, diasporas, or organizations.
These taxonomic views invoke the idea of two bounded
spaces: “here” and “there,” where “there” is other, apart,
and disconnected, stably distanced from “here.” They
invoke “other-ness” and, tacitly, a universal “self” who can
observe and mark the difference. Even within the “here”-
ness of the American town that was the researcher’s field
site, the researcher was promised to encounter an “other”
“out there” – someone with a different, disconnected, and
static method of categorizing and making sense of the
world.
The promise of generalizability along familiar scales such
as the nation-state has made Hall’s [17] and Hofstede’s [19]
frameworks popular in organizational behavior and social
informatics. The taxonomic view, however, suffers from a
range of problems. Studies have found such dimensions
analytically weak in explaining conflicts [10] and
differences in technology use [12], while HCI researchers
have argued that cultural “averages” are of limited use for
design [24]. More problematically, the taxonomic view of
culture is both historically and ontologically suspect.
Cultural categories are frequently rooted in geographical
separation, but technologically enabled interactions, such as
internet communities, mobile technologies, and remote
collaboration, call into question where one cultural zone
ends and the next one begins. These technologies of
communication and mobility circulate cultural concepts
globally ([3,22]) as a condition of contemporary (and not-
so-contemporary) living. At what scale should we see
culture – the nation, the region, the city, the town? With
respect to the “Asian”, “European” and “Latino” families
encountered in this fieldwork, where does the “home
culture” end and the “American culture” start?
A more productive analytical position here, we argue, a
“generative” view of culture arising from contemporary
anthropology and postcolonial studies. Here, culture is a
lens through which people collectively encounter the world
– a system of interpretive signification through which the
world inter-subjectively meaningful. From this view, an
individual may participate in many cultures – cultures of
ethnicity, nationhood, profession, class, gender, kinship,
and history – each of which, with its logics and narratives,
frames the experience of everyday life. Rather than
classifying people on various cultural dimensions, a
generative view of culture suggests we ask how the
technological objects and knowledge practices of everyday
life become meaningful contingently and dynamically as
social activity unfolds. In this sense, culture shapes
experience but is in turn reproduced and generated through
everyday interaction.
Indeed, taking culture as something that is dynamic,
collectively produced, and enacted in everyday encounters
problematizes taxonomic models that see culture as
acquired and internal to the individual – “software for the
mind” shared by people of the same nation [19]. This shift
of perspective is especially important in the context of HCI,
as understanding transformations effected in part by
technology design requires an understanding of cultural
change as much as cultural stabilities. Hofstede’s popular
framework, after all, provides a snapshot of traits at a single
point in time. It has little to say, then, about the norm-
shifting of technologies, social movements, or even
everyday reconfigurations of practice around technology,
media, artifacts or experience. Yet it is precisely those
changing cultural practices that designers aspire to support
and in which they wish to intervene when they introduce a
system into a setting. For example, HCI4D projects might
use technology to make microlending more efficient,
accountable, and far reaching (e.g. [26]). Such interventions
begin in conversation with existing practices, but also
reconfigure local power relations within villages and
households [16]. These reconfigurations can shift the
interpretive frames of diverse aspects of everyday life,
including technology, financial practice, commercial
activity, and gender relations. Understanding how
technology design is adopted, learned, and used, then,
requires a dynamic model of culture.
We do not present this case to suggest that studies of
diasporic communities can substitute for studies of cultures
in a pure or home context. Rather, the diaspora case
illuminates the fluidity of cultural, regional, and
transnational boundaries, as well as the variability of the
what “home culture” can mean in daily life. People relate
variably to a range of local and international networks,
producing their cultural identities through a variety of
signifying practices [20]. If it seems that Indonesians, say,
share common frames in how they intersubjectively
interpret the world, it is a phenomenon in need of
explanation. The same may not be true of all nations.
Consistent with these dynamic processes by which cultural
identities and practice are formed, the generative view of
culture suggested by the postcolonial perspective allows
designers to recognize their work not as designing
appropriately for static, nationally-bound cultures, but
instead as interventions both in conversation with and
transformative of existing cultural practices.
The Problem of Development
Our second case is drawn from an initial ethnographic
engagement that forms part of a larger, ongoing project.
D-Design, an Indian design consultancy, had been
commissioned by health care NGO HealthWorks to
complete a home placement study of prototype household
water filters. The client was a reputable non-governmental
organization focused on development. HealthWorks sought
to develop a commercial market for water filters among
Indians living in poverty as a means of curbing water-borne
illness. The imagined study recruit, according to the lead
designer, was “fairly poor,” getting “water from the dirty
river,” often ill from water-borne illness, and without a
filter.
D-Design’s team drove hundreds of kilometers from village
to village searching for participants for the study but found
few people who matched the client’s image of poverty. What
the design researchers found instead were villages where
people seemed relatively happy or even proud of their
water. Few complained about water-related illness, though
many complained about health problems from over-
fluoridated water from wells – a problem that the prototype
water filters would not remedy and HealthWorks chose not
to pursue. “Where is the poverty?!” cried one of the
designers at a team debriefing following village visits,
throwing his hands up before dramatically throwing his
head onto the table. Failing to find the imagined targets of
development in the field, the team loosened their image of
the ideal participant, finding people who were curious
enough about the filter and met loosened income
requirements.
By seeing this case through the analytical lens of
postcolonial computing, several issues in contemporary
development are foregrounded: the discourse of global,
technical solutions to problems; the alignment of
development projects with the interests of commercial
actors in industrialized countries; the directionality of
product and monetary flows in development programs; and
disempowerment through consumer-oriented development.
Development here encompasses a range of programs for
financial and technological assistance set up between
“advanced” and “developing” countries, or among more
complex financial arrangements between governments,
NGOs, philanthropists, corporations, and supranational
institutions such as the World Bank.
The discourse of global, technical solutions to problems of
development is critical to understanding why HealthWorks
chose not to pursue the problem of over-fluoridation
identified through early stage design research. This may
seem to fly in the face of user-centered design, which
prescribes designing around target users’ needs. The
project, instead, is a solution in search of users – users
found through great expenditure of time and energy.
HealthWorks’ reluctance to change course stems from the
wide availability of low-cost bacterial filtering technology
designed for wealthier markets. HealthWorks, by their own
description, adapts existing technologies in rich countries so
they work in poor ones. Designing a fluoride filter would
require a longer design and research cycle for a less
universal-seeming problem. By contrast to the NGO’s
design-centered efforts, local Indian activists have been
mobilizing for a public-sector, political and infrastructural
solution to the fluoride problem. Ferguson’s analysis of
NGOs and development in Lesotho [13] suggest that
HealthWorks’ global framing is characteristic of
development more broadly; development regimes in recent
decades, he argues, have systematically avoided
confronting the actions of large-scale actors such as
governments and corporations as causes of the socio-
economic conditions they seek to remedy, instead seeking
behavioral, educational, and market-based solutions at the
local level. Some see NGO’s focus on the local as an
advantage, keeping NGOs far from the taint of the politics
[15] of governments that have been painted as corrupt. This
NGO avoidance of politics, Ferguson argues, depoliticize
poverty by covering the sometimes highly political causes
of poverty: “By uncompromisingly reducing poverty to a
technical problem, and by promising technical solutions to
the sufferings of powerless and oppressed people, the
hegemonic problematic of ‘development’ is the principle
means through which the question of poverty is de-
politicized in the world today” [13:256]. In Lesotho,
Ferguson’s example, World Bank reports recommended
agricultural technology as a remedy for farmers who had
lost their arable land to Dutch settlers. These approaches
fail to discern the importance of political causes and
political solutions to problems of poverty, leaving many
failed development projects in their wake.
This leads to a second critique: development regimes have
historically been aligned with the interests of politically
powerful commercial and capital market actors.
HealthWorks is well known as an early advocate of public-
private partnerships in the development and distribution of
health products. In its project with D-Design, HealthWorks
sought not only to develop a usable, useful water filter but
also to develop distribution networks to retail the filter
through rural and slum businesses. In aligning diverse
actors to enable the distribution of water filters,
HealthWorks’ interests are aligned with corporations like
Pepsi who have similarly developed retail channels among
India’s poor. These alignments can take diverse forms at
different times and places. In opening developing
economies to Western investment as a loan condition,
World Bank policies destabilized many economies by
exposing them to speculation or predatory corporate
interest. In some cases, these interests have not run parallel
to those of stable state development. In his study of
extractive industries in Africa, Ferguson [14] notes that
while opening African nations up to Western investment in
the 1990s had been proposed as a way of incorporating
African nations into the global economy and providing
them with the resources to grow both economic and civic
infrastructure, foreign investment actually fared best in
countries that were anti-democratic or even in a state of
civil war. Highly-technologized mineral extraction
industries reaped profits by avoiding African nations with
strong unions and democratic traditions as they would
impede extraction operations. The potential consequences
of bringing resources and people in line with the interests of
powerful capital and commercial actors is rarely recognized
and analyzed in ICT4D literatures.
Third, the flow of technologies from HealthWorks, an
American NGO, to development targets around the world
parallels the structural flows of goods and money in
colonial relations. The economic ties are quite complex.
HealthWorks partners with private entities, sometimes local
distributors or manufacturers and at other times
multinational consumer goods corporations, to develop and
distribute health products. As a prominent, well-funded
NGO with connections to powerful actors, however,
HealthWorks exerts considerable strategic control over the
shape of projects and deployments by channeling money in
ways that funders in highly industrialized countries find
compelling. These flows, Escobar argues [11], of capital
and strategic control characterize development more
broadly. Escobar analyzes the history of development
efforts in the second half of the twentieth century,
observing that development typically follows patterns based
broadly on colonial histories. The economics of colonialism
were frequently based on the movement of raw materials
from colonies to colonial centers, and the movement of
finished products in the opposite direction. The movements
of goods in development relationships are often much the
same.
Fourth, HealthWorks’ local, individually-focused strategy
attempts to build a market of consumers – a strategy that
some argue is inherently disempowering. Escobar argues
that development programs have historically pursued
progress by figuring people as consumers of technology
designed and manufactured elsewhere. These technologies
historically have included laptops made in Taiwan [6],
fertilizers engineered in the US [16], or even water filters.
In Escobar’s analysis, the rhetoric and practice of
development positions emerging nations as essentially
powerless and unable to “develop” without intervention.
The provision of goods and services then positions these
nations as consumers of Western technologies and services.
The kinds of infrastructure investments that are often
associated with development efforts ensure “lock-in” to the
goods and services provided by Western corporations. This
was an explicit goal of the US Marshall Plan for aid to
Western Europe after the Second World War, which sought
to rebuild the economies of European nations in order to
ensure a ready market for American products [30]. The
Marshall Plan set the context for further development
programs, which share both the ideological commitment
and the practical consequence of creating new markets and
new consumers. These framings do not necessarily effect
disempowerment, but they make uneven relations of
economic dependency possible. These problems complicate
HCI4D projects that seek to help integrate people into
markets, whether as producers or consumers, as a means of
empowerment.
Uneven Economic Relations
No design practice takes place outside of a series of
economic conditions that make it possible. In the case that
follows, uneven economic relations, with consequences for
legal and cultural power, shape what counts as legitimate
design practice.
As Brazil sought to develop its economy during the 1970s
and 1980s, its government provided incentives to
companies that designed and manufactured products within
Brazil as a means of eliminating poverty and encouraging
indigenous innovation. Unitron, a Brazilian company,
responded by reverse engineering, designing, and
manufacturing a Mac compatible machine called “Mac de
periferia” (Mac of the Periphery). Though Apple had no
intellectual property protection for the Macintosh in Brazil,
the American corporation was able to pressure government
and other economic actors within Brazil to reframe
Unitron’s activities, once seen as nationalist and anti-
colonial, as immoral piracy. In exerting political pressure
through its economic strength, Apple was able to reshape
notions of authorship to exclude reverse engineering and
modification, realigning Brazilian notions of intellectual
authorship with American notions that privilege the
originator over maintainers and modifiers of code and
hardware architecture [9]. These economically motivated
legal actions shaped definitions of what counts as
legitimate design work, innovation, and creativity –
concepts often taken for granted in HCI research.
Notably, in this case, Unitron’s design process had value
beyond the technology that it produced, or in this case
reproduced. The very process of creating the Mac of the
periphery was meaningful as a claim of indigenous
technological capability in reverse engineering, as a symbol
of local manufacture, and as a way of signaling national
independence. This suggests that as design methods travel
transnationally, we consider not only how requirements are
gathered or how artifacts are produced, but also by whom,
in what economic relationship, and with what cultural
meanings. In contrast to HCI culture, which commonly
values creativity and the production of the new [34],
Unitron represented the value of reconfigured economic
and technological relations in Brazil during the 1970s –
relations then reshaped by more powerful economic actors
within the global industrial system.
The cultural meanings of design practices are also evident
in the adoption of open source technologies in Peru. In the
early 2000s, the Peruvian government and open source
activists began arguing for free software (FS) over
proprietary software for official uses. Activists first argued
that FS reduced dependence on transnational corporations
and later argued that FS ensured public control over
information access [7]. Peruvian open source at the time
became not only an appropriation and translation of
software design methods of knowledge and software
production that originated in the West, but also a response
to contemporary conditions surrounding intellectual
property and the international economics [9,27]. It was not
only the usability, learnability, or even fun of such
technologies that shaped their adoption and diffusion.
Technologies’ cultural meanings in particular social and
historical circumstances, not only among users but among
others [29], that shaped their uptake.
These cases suggest that the uneven flows and exchanges of
capital that surround the uptake of information technologies
have implications for HCI interests such as design method,
software adoption, usefulness, usability.
Knowledge and Voice
A Skype call, in February 2009, connects ten researchers
and collaborators across the world. The context for the
call is a project, funded by the Australian government,
investigating the use of Internet conferencing technologies
to allow indigenous Australian knowledge authorities to
teach university classes about their culture and language
from their homelands. Participants in the call are
connected from California, Darwin, Melbourne, and a
number of sites in the traditional homelands of the Yolngu
people in Northeast Arhemland. In reference to a recent
lecture in which one of the indigenous participants had
discussed a story – the Turtle dreaming – one of the
researchers, H, suggests to the knowledge authority, D, that
perhaps an interesting class would consist of a discussion
of this story amongst indigenous participants. H asks,
“Does your brother still live next door? Perhaps you and
he could talk about it together on camera.” D is open to
this, but immediately reluctant. The Yolngu observe a
brother/sister avoidance relationship, but there is another
problem. Speaking of her current location, she says, “These
are our mari’s [classificatory grandmother’s] lands. We
don’t have the authority to tell the Turtle dreaming here.
We could do it if we went back up to [home settlement]. Ah,
but we don’t get good 3G service there. We’d have to use
the satellite phone; and the delay on that [for
videoconferencing] is always a problem.”
Our traditional design methods are centrally concerned with
the problem of making knowledge portable. Knowledge
about usage problems and settings must be moved from the
context in which it arises (the users’ everyday world) to the
design site; knowledge about design opportunities must be
moved from the design site to the development process;
knowledge about the design must somehow be moved
among those participating in the creation of the technology.
Even when these movements are seemingly simple – from
the meeting room to the cubicle – a whole range of
representations and artifacts are employed to achieve them.
More significantly, we depend on an infrastructure of
knowledge practices to achieve these movements – the
practices that allow us to interpret the artifacts, to invest
them with authority, to recognize the people who stand
behind them, and so on. As the case of the Yolngu shows,
knowledge sharing – what it means to know something, and
what it means to be able to tell it – is hemmed in all around
by a series of infrastructures, social, cultural, and
technological, that must be brought into alignment.
Clearly, this affects any efforts in “knowledge
management” by which diverse cultural understandings are
to be harnessed to western techno-social practice (e.g.
[4,35]). In recent work, Verran and colleagues have
attempted to understand these questions specifically in the
context of online practice and digital information, with the
goal of developing an indigenous aboriginal approach to
information management [35]. Indigenous knowledge, we
would caution, is not to be naively contrasted with scientific
here. Verran’s project, and technologies more generally,
occasion new hybrid forms (e.g. the hybrid chemical-
humoral models of soil that Gupta notes in his studies of
agricultural development in rural India [16]). The
“knowledge” that flows around these design processes is
not impersonal, abstract, or timeless. It emerges at
particular moments, voiced by particular people, who speak
with particular kinds of authority – cultural, technical,
spiritual – and their right to speak is contingent on that
authority. Srinivasan’s [32] projects similarly attempt to
create information systems with community-appropriate
ways to categorize and organize information and artifacts.
His work emphasizes the importance of specific community
“voice” in contrast to the (implicitly unvoiced) notion of
abstract authority in data repositories.
The cultural conditions of practices of knowing and telling
become deeply relevant to HCI design methods because
HCI methods themselves embody principles and models of
knowledge sharing and knowledge representation.
INTERCULTURAL ENCOUNTERS IN DESIGN PRACTICE
As we have illustrated, the insights of postcolonial studies
and STS speak at once to the highly local and contingent
practices that we see at work in different specific sites of
technology design and use, while at the same time
recognizing the ways that those localisms are conditioned
and embedded within global and historical flows of
material, people, capital, knowledge, and technology. The
postcolonial computing lens focuses our attention, then, on
the translations, dependencies, conditions, and histories that
shape perceptions of technology and its opportunities. This
change in approach inspired by postcolonial and STS work
provoked us as designers to revisit our design toolkit with
an eye towards international settings.
This reframing also invites us to reconsider design methods
and design practice. Traditional design processes, in HCI as
well as related disciplines, break down into a familiar range
of steps and procedures, from the identification of potential
user communities, the analysis of their activities, the
formulation of design requirements, ideation and iteration,
and so forth. We suggest an alternate formulation of design
work – engagement, articulation, and translation. This
formulation broadens notions of what counts as design
work and suggests an alternate sensibility for evaluating
such work.
By “engagement” we mean connecting with users or an
application domain in order to understand relevant work or
activity; since HCI design is primarily “user-centered,” this
is most commonly (although not always) a form of
engagement with people and their material worlds.
“Articulation” concerns how properties of this domain are
formalized and transformed into a series of requirements for
technological support. Finally, “translation” concerns how
these requirements, possibly through a series of steps, are
transformed from statements about a domain to statements
about technology and eventually into specific pieces of
technology designed to support the application domain.
By abstracting design methods into these aspects, we gain
analytical purchase over HCI design practices. Our
motivation for introducing these facets is to take traditional
elements of the design process, such as requirements
capture or workflow analysis, and place them in a context
that makes clear relevance of issues of power, history, and
epistemology. We will discuss each one in turn, and show
the relevance of a postcolonial approach.
Engagement
Postcolonial computing points to the many ways histories,
power relations, and epistemology tacitly underpin
engagements in design, offering HCI researchers and
practitioners new lines of inquiry. Design often involves
transcultural encounters, between designers and users,
designers and engineers, between corporate- and non-
corporate actors, or even between differently located actors
working collaboratively to design for themselves or for one
another. In this view, observing users, eliciting
requirements, and deploying technologies are only some of
the sorts of encounters that can animate HCI research.
Engagements in design methods can take many forms,
including design research, technology deployment,
feedback, and maintenance. In usability testing, people
serve as model users in a usability lab setting, directed by a
usability researcher to test software in front of two-way
mirror. The encounter within the laboratory walls is
described as data, often observed by unseen others behind a
one-way mirror or on video. Whether or not this form of
engagement is scientific or objective is less interesting for
our purposes than the way the engagement has been shaped
by scientific roles and discourses, lending authority both
between researcher and model user, and between researcher
and consumer of the research findings. The authority of
this method benefits from the ways scientific ways of
knowing have been elevated above other forms of
knowledge creation in some settings [1]. In other settings,
however, the context in which usability studies work may
not be present. For example, histories of exploitation by
researchers or a distrust of disengaged observers make such
methods untenable in some communities [4].
Another Western design method, Participatory Design
(PD), draws authority from a very different set of
discourses of labor relations and social justice. Developed
primarily in Scandinavia but taken up more broadly, PD
maintains roles for designers and users but calls for users to
participate in the imagination and specification of work
technologies. The modes of engagement, deliberation, and
negotiation are drawn from Scandinavia’s strong traditions
of union involvement in workplace decision-making [36]. It
is not that PD works because of an inherent superiority to
other methods. It works (when it works) because it takes
advantage of cultural logics and practices particular to the
location in which it emerged. Researchers have long
grappled with how to generalize and adapt PD to enable
participation by non-organized workforces and have noted
the particular challenges of reproducing PD engagements in
different national and political cultures [25].
Drawing on Suchman’s [33] account, we find it useful to
think of design processes less as ways that designers can
formulate needs and measure outcomes, and more as
shaping and staging encounters between multiple parties.
The essence of the process is the fact of different people
coming together and meeting – holding a conversation
rather than following a recipe. What this draws our
attention to, then, is, first, the context of that encounter – its
histories, its politics, those who are present and those who
are not, the authorities and responsibilities with which
different people speak – and, second, the bidirectionality of
the exchange. It directs us to think about what people bring
into the encounter and what they take away from it.
For instance, consider Hayden’s [18] discussion of
bioprospecting processes in Mexico, in which scientists
from US and European pharmaceutical companies sought to
identify the active ingredients in traditional remedies that
could then be incorporated into new drugs. This encounter
is framed by very different understandings of the kind of
knowledge being shared – one of tradition and spiritual
practice and one of intellectual property licensing. The
encounter breaks down not least in light of these different
orientations, which similarly inflect any understanding of
what is being “exchanged.”
There are two consequences to reframing design methods
from extractive processes, such as lessons learned,
knowledge gained, or requirements identified, to mutual
encounter and learning in which responsibilities between
different parties are enacted. First, it suggests we
acknowledge users as active participants and partners rather
than as passive repositories of “lore” to be mined. People
may bring different things to the table and may take
different things away; indeed, they might have different
understandings of what has been brought and taken. A
focus on the encounter as a moment of mutual engagement
gives us a very different way of thinking about what
“requirements”, say, are, and what the process of “capture”
might entail. The second lesson is the recognition of an
encounter as an intentional, motivated, and power-laden act
(rather than as an inherent consequence of an impersonal
process.) People are brought to the table; others are
relegated to peripheral positions or excluded altogether.
This is always and obviously the case, and indeed it is
frequently desirable (some people speak up when others are
out of the room); the point is to be conscious of it and to
imagine alternatives.
Articulation
By articulation we mean the way we might see and frame a
situation as designers. In much design practice, articulation
includes the ontological work of designating targets of
design, such as an individual or a community, and
interpreting or inferring the targets’ “needs, “wants,”
“desires,” “opportunities,” or “constraints.” Articulation
abstracts the experiences of engagement and reframes them
in terms amenable to design practice. Articulations are
culturally specific for both designers and their audiences.
The production of formal and informal representations is a
key component of articulation, including personas that
serve as stand-ins for imagined users, task flows that
represent work processes in designed-for contexts, or
research summaries that conclude with implications for
design. These representations are normally produced to
move around in the world – to flow from designers to
implementers, to document decision-making processes, to
serve as a baseline for later evaluation, and to be reused and
repurposed in future design efforts. As our discussion of
knowledge practices suggested, however, a postcolonial
perspective might point to the highly situated nature of
knowledge practices, with two implications for HCI4D.
First, HCI methods often see knowledge as something to be
captured, rather than something to be performed or enacted
as it is in other cultural settings. The photograph only
captures something of the moment; the question of whether
it captures the essence depends on what you think is
important about the moment. Second, by corollary, it these
seemingly mobile representations are themselves highly
localized, connected to specific practices of seeing and
interpretation. Latour [23] calls such representations
“immutable mobiles”, but they are “immutable” only within
delimited ranges of movement. Representations of people,
activities and artifacts take on radically different meanings
in different cultural contexts in ways that can often be quite
problematic. Only in the context of a specific set of
technical practices and assumptions does a workflow
diagram capture what it means to carry out a task. Cross-
cultural design efforts, then, must recognize how
articulation in UCD functions to frame and interpret
engagement drawing on traditions of representation that
may not be commensurable with knowledge and design
practice in other cultures. Processes of articulation come
with ontological and epistemological commitments.
Furthermore, perfect commensurability through translation
is not possible. Postcolonial computing is not a matter of
finding the right ethnographic informant or the true way of
articulating users’ ontologies. It is a matter of grappling, as
did the designers of the aboriginal knowledge system
discussed earlier [35], with how to design when the
certainty of perfect intercultural translation is not possible.
We can see these issues at work in the OLPC project. The
project’s motivating articulation of needs is made with
respect to a very familiar, unilinear model of technological
progress, one in which the problems of the developing
world are framed as a series of absences, specifically of the
furniture of Western life, viz., digital technology, and in
particular, the personal computer. Implicit in this model are
commitments to individual ownership; indeed, one of the
critiques of OLPC has been that the devices might have
been better conceived as owned by communities or family
groups, and yet the specific design of the laptop – and in
particular the small keys on the keyboard, or even its color
and design – orient it specifically towards children rather
than other family members. Articulated design needs and
requirements, in this case individual control and ownership,
reflect a series of often unquestioned ontological, political,
and economic commitments that are integral to the design.
Translation
Finally, how might this postcolonial perspective illuminate
“translation” in design methods: the transformation of
requirements into statements about technology or
technology itself? Translation has both a linguistic sense
and a geometric sense. The linguistic sense of “translation”
captures the transformation between different languages –
representational schemes that are deeply culturally situated.
At the same time, “translation” also has a geometric sense,
where it refers to the movement of a figure from place to
place. In the examples we are discussing, both are present
concurrently. Although the very idea of design methods as
portable prescriptions of practice is predicated on the
assumption that translation preserves meaning, a focus on
transformation and movement may allow us to be more
attentive to the fact that such translations are intentional and
explicit acts whereby statements produced for one purpose
are made amenable to new purposes and new actors. The
translations, then, by which statements about domains and
needs become statements about technology and solutions,
must also be seen as purposeful, partial, and situated.
While HCI research has recognized the cultural specificity
of design products, the processes and methods of design has
largely been imagined as universal. Taking a broader view,
we would argue that methods – the products of research
communities, economic actors, and educational practices
that span the globe – are always transnationally produced
and dynamic. When an open source software (OSS) team or
a proprietary design team meets in Silicon Valley, they are
simultaneously negotiating their products and their methods
in situ. The two are not easily separable. In the same way,
Indian OSS teams or proprietary software designers are also
engaging in the same talk and practice around their
products and methods. As design methods move around in
the world, they are subject to different interpretations and
they become locally meaningful design in different ways.
Engagement, articulation, and translation are only three
aspects of varied, highly situated practices of design. Yet
examining these aspects through the lenses provided by
STS and postcolonial studies exposes some tacit
commitments and thus limitations of current approaches.
CONCLUSIONS
We have raised a number of issues relevant to information
technology drawn from STS and postcolonial studies. We
have argued that these considerations are particularly
significant in the context of current migrations HCI into
new settings, but they have been largely absent from the
research literatures in this area. Yet we do not mean to
suggest that design over “there” is fundamentally different
than design “here.” While the cultural traditions of UCD
have often emerged out of Western traditions, those
traditions themselves are contradictory and multiple. There
is no single “here” of Western knowledge and design.
Incommensurability might mark relations not only between
different national cultures, but also organizational cultures,
or even professional communities within an organization.
Our goal is not to criticize ongoing work in HCI4D and
cross-cultural design, but rather to expand the conversation
around cross-cultural technology development by placing it
in a broader context. Understanding the processes of
translation of design methods (both in the sense of
movement and in the sense of transformation) is a means to
this end. Similarly, it is important to avoid simple dualisms
between developed and developing contexts, traditional and
scientific knowledge practices, and so on. In his study of
development efforts, Gupta shows that simple dualities of
traditional and modern, developed and undeveloped,
colonial and (historically) postcolonial do not hold [16].
Rather, we call for an investigation of how, out of the
encounter between different forms of knowledge and
practice, new hybrid forms design and technology are
generated. Communication is bidirectional here, and as the
postcolonial perspective vividly demonstrates, the
encounters here are not instantaneous and fleeting, but
rather take place over years and decades.
We argue for attentiveness to the emergence of hybrid
practices in information technology design, coupled with
sensitivity to how uneven power relations are enacted in
design practice. The goal, as we have outlined here, is not
simply to bemoan the problems that arise when methods are
assumed to move easily and stably from one setting to
another, but rather to understand the diverse forms design
practice and contextual reasons for that diversity. We have
suggested that thinking about the design process in terms of
engagements between different groups, the complexities of
articulating perspectives, and the implications of translation
between sites, provides a starting point for acknowledging
and embracing heterogeneity in design, rather than
attempting to control or eliminate it.
While we have been looking at design primarily in the
context of HCI4D, concerns of power, mutual intelligibility,
and how cultural forms are generated are also relevant to
the intercultural encounters between designers and users in
traditional corporate or academic contexts. Suchman [33]
framed the work that she and her colleagues conducted at
Xerox as the staging of encounters amongst various
stakeholders, including engineers and workers. The lens of
HCI4D defamiliarizes design so that we may understand its
methods more deeply. Locating these commitments allows
us to imagine what practices with alternate commitments
might look like, and open new spaces for design [33].
ACKNOWLEDGMENTS
We thank Geof Bowker, Leigh Star, Silvia Lindtner, Helen
Verran and our anonymous reviewers for their close
readings and feedback on this work. This work is supported
by NSF Graduate Research Fellowship, Intel PAPR@UCI,
and NSF grants 0712890, 0838601, 0838499, and 0917401.
REFERENCES
1. Adas, M. Machines as the Measure of Men: Science,
Technology, and Ideologies of Western Dominance.
Cornell University Press, Ithaca, NY, 1989.
2. Akrich, M. The De-scription of Technical Objects. In
Shaping Technology / Building Society. MIT Press,
1992.
3. Appadurai, A. Modernity at Large: Cultural
Dimensions of Globalization. University of Minnesota
Press, 1996.
4. Bidwell, N.J., Standley, P., George, T., and Steffensen,
V. The landscape's apprentice: lessons for place-
centred design from grounding documentary. Proc.
DIS 2008 (Cape Town, South Africa), ACM Press, 88-
98.
5. Brewer, E., Demmer, M., Du, B., et al. The case for 341.
technology in developing regions. Computer 38, 6
21. Marcus, A. and Gould, E.W. Crosscurrents: cultural
(2005), 25-38.
dimensions and global Web user-interface design.
6. Butler, D. The race to wire up the poor. Nature 447, interactions 7, 4 (2000), 32-46.
7140 (2007), 6-7.
22. Marcus, G.E. Ethnography in/of the World System:
7. Chan, A. Coding Free Software, Coding Free States: The Emergence of Multi-Sited Ethnography. Annual
Free Software Legislation and the Politics of Code in Review of Anthropology 24, (1995), 95-117.
Peru. Anthropological Quarterly 77, 3 (2004), 531-
23. Marsden, G. Designing technology for the developing
545.
world. interactions 13, 2 (2006), 39-59.
8. Chetty, M., Tucker, W., and Blake, E. Developing
24. Marsden, G., Maunder, A., and Parker, M. People are
locally relevant software applications for rural areas: a
people, but technology is not technology.
South African example. Proc. SAICSIT 2004
Philosophical Transactions of the Royal Society 366,
(Stellenbosch, South Africa), ACM Press, 239-243.
(2008), 3795-3804.
9. da Costa Marques, I. Cloning computers: From rights
25. Muller, M.J. Participatory Design: The Third Space in
of possession to rights of creation. Science as Culture
HCI. Handbook of HCI. Erlbaum, (2003).
14, 2 (2005), 139.
26. Parikh, T.S., Javid, P., Ghosh, K., and Toyama, K.
10. Easterbrook, S.M., Beck, E.E., Goodlet, J.S.,
Mobile phones and paper documents: Evaluating a new
Plowman, L., Sharples, M., and Wood, C.C. A Survey
approach for capturing microfinance data in rural
of Empirical Studies of Conflict. CSCW: Cooperation
India. Proc. CHI 2006 (New York, NY), ACM Press,
or Conflict, (1993), 1-68.
551-560.
11. Escobar, A. Encountering development: The making
27. Philip, K. What is a technological author? The pirate
and unmaking of the Third World. Princeton
function and intellectual property. Postcolonial Studies
University Press Princeton, 1995.
8, 2 (2005), 199-218.
12. Ess, C. and Sudweeks, F. Culture and Computer-
28. Ramachandran, D., Kam, M., Chiu, J., Canny, J., and
Mediated Communication: Toward New
Frankel, J.F. Social dynamics of early stage co-design
Understandings. Journal of Computer-Mediated
in developing regions. Proc. CHI 2007 (San Jose CA),
Communication 11, 1 (2005), 179-191.
ACM Press, 1087-1096.
13. Ferguson, J. The anti-politics machine:" development,"
29. Satchell, C. and Dourish, P. Beyond the User: Use and
depoliticization, and bureaucratic power in Lesotho.
Non-Use in HCI. Proc. OzCHI 2007 (Melbourne,
Cambridge University Press, 1990.
Australia, Nov 2009), ACM Press, .
14. Ferguson, J. Governing Extraction: New
30. Schain, M. The Marshall Plan: Fifty Years After.
Spatializations of Order and Disorder in Neoliberal
Palgrave Macmillan, New York, 2001.
Africa. In Global shadows: Africa in the neoliberal
world order. Duke Univ Press, 2006. 31. Smith, A. The Alternative Technology Movement: An
Analysis of its Framing and Negotiation of Technology
15. Fisher, W.F. Doing good? The politics and antipolitics
Development. Human Ecology Review 12, 2 (2005),
of NGO practices. Annual Review of Anthropology 26,
106-119.
1 (1997), 439–464.
32. Srinivasan, R. and Shilton, K. The South Asian web:
16. Gupta, A. Postcolonial Developments: Agriculture in
an emerging community information system in the
the Making of Modern India. Duke Univ Press, 1998.
South Asian diaspora. Proc. PDC 2006 (Trento, Italy),
17. Hall, E.T. Beyond culture. Anchor Books, 1977. ACM Press, 125-133.
18. Hayden, C. When nature goes public. Princeton Univ 33. Suchman, L. Located accountabilities in technology
Press, Princeton, NJ, 2003. production. Scand. J. Inf. Syst. 14, 2 (2002), 91-105.
19. Hofstede, G. and Hofstede, G. Cultures and 34. Suchman, L. Striking Likeness to Difference.
Organizations: Software for the Mind. McGraw-Hill, 4S/EASST, (2008).
2005.
35. Verran, H. and Christie, M. Digital Technologies and
20. Lindtner, S., Mainwaring, S., Dourish, P., and Wang, Aboriginal Knowledge Practice. (2006).
Y. Situating Productive Play: Online Gaming Practices
36. Winograd, T. Bringing design to software. ACM, New
and Guanxi in China. In INTERACT 2009. 2009, 328-
York, NY, USA, 1996.